Deep Learning was introduced about 30 years ago to exploit the concepts related to neural networks, derived from Artificial Intelligence. This branch of machine learning has recently undergone the most significant advances, making it particularly effective in learning contexts underpinned by non-linear relationships.
This training is aimed at data scientists and developers who already have a minimum of knowledge and practices in machine learning: the basic concepts will be reviewed as an introductory part. The main goal is to focus on concrete manipulations of TensorFlow. Several functional themes will be applied to it, especially those highly media-based images and text recognition.
The training will present, in a relatively detailed form, the most interesting algorithms proposed by TensorFlow. Some very useful features, allowing for example the dynamic visualization of data with TensorBoard or the putting into production of the models built thanks to TensorServing, will also be studied.
- Understand neural networks
- Understanding Deep Learning
- Develop models with TensorFlow
- Knowledge in Machine Learning
- Basic knowledge of algebra (matrices) and statistics
- Knowledge of Python programming
Developers, Data Scientists, Architects
Financing in France
- May be financed through OPCA (if financing covers all of the cost of the training)
- Cannot be financed through the CPF